Drug-Drug Interaction Detection: A New Approach Based on Maximal Frequent Sequences

In this paper, a new approach for Drug-Drug Interaction detection is presented. The proposed method consists in discovering patterns automatically through Maximal Frequent Sequences extraction and using pattern matching to identify sentences that contain Drug-Drug Interactions. Maximal Frequent Sequences define word sequences that are frequent in texts and it has been proved in this paper to be a good method for DDI detection, obtaining promising results with high values of precision and recall. The method proposed is domain and language independent.